BJA Advance Access published online on June 20, 2008
British Journal of Anaesthesia, doi:10.1093/bja/aen183
An unusual explanation for low oxygen saturation
1 Department of Anaesthesia, St James’s University Hospital, Beckett Street, Leeds LS9 7TF, UK
2 Department of Anaesthesia, The General Infirmary at Leeds, Great George Street, Leeds LS1 3EX, UK
* Corresponding author. E-mail: s.p.holbrook{at}leeds.ac.uk, simonholbrook{at}btinternet.com
Accepted for publication May 29, 2008.
 |
Abstract |
|---|
 Top Abstract IntroductionCase reportResultsDiscussionReferences |
|---|
Pulse oximeters are now commonplace in modern medical practice, but we still need to be aware of their limitations. We present here a case of a 62-yr-old gentleman who underwent general anaesthesia for a recurrent parietal meningioma. He had received multiple general anaesthetics in the past. Persistent low pulse oximetry readings in the perioperative period, without any suggestion of respiratory compromise, led us to investigate him further and to discover a new, but benign, haemoglobinopathy.
Keywords: blood, haemoglobin; complications, oxygen desaturation; equipment, pulse oximeters; measurement techniques, oximeters
|
Introduction |
|---|
|
TopAbstractIntroductionCase reportResultsDiscussionReferences |
|---|
The first pulse oximeter became commercially available in 1975.1 Anaesthetists soon realized the advantages to be gained through the use of the pulse oximeter: primarily as a patient monitoring tool and an aid to improving patient safety. Their use is now considered essential in modern anaesthetic practice.2 We report here a case of a patient undergoing general anaesthesia whose pulse oximetry readings were persistently low without any evidence to suggest respiratory pathology.
|
Case report |
|---|
|
TopAbstractIntroductionCase reportResultsDiscussionReferences |
|---|
A 62-yr-old man presented for elective surgery for excision of a recurrent right parietal meningioma. The diagnosis was made in 1997 and a primary excision was performed. However, he underwent three further operations as a result of surgical complications: a total of four general anaesthetics in a 15 month period. The patient reported no anaesthetic complications during these. He was an ex-smoker with a history of well-controlled, mild, seasonal asthma. He denied any symptoms on enquiry of cardiovascular and respiratory systems and clinical examination was completely unremarkable. He was not cyanosed. A chest radiograph was normal as were all laboratory investigations.
In the anaesthetic room, standard monitoring was applied. On room air, the SpO2 recorded a value of 90%: this value increased to 93% when oxygen was administered at 6 litre min–1 via an MC facemask. The plethysmograph trace was of good quality and the values did not improve with either alteration of the probe site or change of the probe itself. The pulse oximeter had worked satisfactorily on another patient before this case and recorded normal values from members of the anaesthetic team. At this point, we reviewed the patient’s case notes, incorporating previous anaesthetic charts. The records stated the same puzzling problem with these comments: ‘referral to a respiratory physician may be worthwhile’ and ‘needs CXR’. The general anaesthesia had nevertheless proceeded otherwise uneventfully on all previous occasions and without any suggestion of detriment to the patient.
The decision was made to proceed with the procedure: the dilemma and anaesthetic technique were fully explained to the patient. A 20 gauge radial arterial cannula was inserted preinduction. Arterial blood gases were measured before induction and revealed the PaO2 to be 11.79 kPa (Table 1). The patient was induced i.v. with a remifentanil infusion and bolus doses of propofol and atracurium, after preoxygenation. The airway was secured with a size 9.0 cuffed reinforced oral endotracheal tube. Anaesthesia was maintained with a sevoflurane/air/oxygen mixture via a circle breathing system, in addition to the remifentanil infusion. The lungs were ventilated with a Datex Ohmeda AS/5 anaesthetic machine in pressure control mode.
|
The operation progressed for 3 h in total. Arterial blood samples were obtained throughout at varying fractional inspired concentrations of oxygen (Table 1). The results are tabulated with the corresponding pulse oximetry readings.
As the case proceeded, we began to suspect an undiagnosed haemoglobinopathy and we discussed this with a haematology specialist registrar. Methaemoglobin levels peaked at 0.7% during anaesthesia, well within the normal range of 0–1.5%. There was no laboratory evidence of haemolysis. Following discussion with the patient and informed consent, mass spectrometry was performed on the preoperative blood sample. The operation was completed uneventfully and he was discharged home on the third postoperative day before the results being made available.
|
Results |
|---|
|
TopAbstractIntroductionCase reportResultsDiscussionReferences |
|---|
Adult human haemoglobin A has a molecular weight of 64 585 Da. On a simplified level, it consists of two pairs of polypeptide chains: two
and two β chains. Each combines with one haem group. Mass spectrometry undertaken at a regional haematology laboratory revealed that 7% of our patient’s haemoglobin alpha chains were abnormal, as illustrated in Figure 1.
|
The smaller peak at 15 098.21 Da (28 Da lighter than the normal alpha chain) is the abnormal chain. Further analysis involved repeated mass spectrometry of smaller peptide units after further enzymatic digestions. These processes identified that the amino acid valine at position 62 in a normal alpha chain had been substituted with alanine in the abnormal alpha chain. It was subsequently confirmed by the regional haematology laboratory that this amino acid substitution had not been previously described. This was potentially a novel haemoglobinopathy. Unfortunately, we were unable to confirm this: the patient changed his general practitioner and we were unable to obtain further blood samples for DNA analysis for the National Haemoglobinopathy Reference Laboratory.3 In addition, an international database of haemoglobin variants has also confirmed that this amino acid substitution has not previously been described, but that, again, further laboratory analysis would be required to register this as a novel haemoglobin variant.4
|
Discussion |
|---|
|
TopAbstractIntroductionCase reportResultsDiscussionReferences |
|---|
The limitations of pulse oximetry as a result of low perfusion states, motion artifacts, skin pigmentation, and nail polish are well documented.5–8 Pulse oximetry may be utilized safely in patients with thalassaemias and sickle cell disease in the majority of situations, but it may underestimate true arterial oxygen saturation in critically ill sickle cell patients and those suffering a sickling crisis.5 The problems resulting from other haemoglobinopathies are less well recognized.9 Hundreds of structural haemoglobin variants have been described.3 4 10–12 The majority are the result of a single amino acid substitution in one of the globin chains. Structural changes may produce an array of pathophysiological consequences: haemolysis due to instability of the molecule; polycythaemia due to increased oxygen affinity; and cyanosis due to low oxygen affinity or methaemoglobinaemia.10 Methaemoglobin is the most frequent dyshaemoglobin resulting in cyanosis and is well known to confound pulse oximetry as a result of its absorption spectrum: in most cases, this is acquired (e.g. prilocaine); congenital forms are rare.13 Some haemoglobinopathies labelled haemoglobin M (because they resemble methaemoglobin to some degree) have differing absorption spectra leading to inaccurate readings when using pulse oximetry.13 Kuji and colleagues14 have previously described the anaesthetic management of a patient with haemoglobin MIwate. The oxygenation of this dyshaemoglobin (which constitutes 30% of the total) is only 60% at PaO2 of 13.3 kPa. Their patient was centrally cyanosed but asymptomatic. Arterial blood gas analysis revealed a PaO2 of 11.2 kPa when breathing room air, but pulse oximetry was completely unreliable: they also depended upon co-oximetry. The anaesthetic course of their patient, as with ours, was uneventful.
Other haemoglobin variants have been described that have rendered pulse oximetry undependable. Gottschalk and Silverberg15 reported a patient with haemoglobin Köln. HbKöln itself actually exhibits increased oxygen affinity and is unstable resulting in haemolysis. Pulse oximetry was inaccurate: SpO2 89% on room air, PaO2 13.5 kPa from an arterial blood gas sample. Methaemoglobin levels were 3.7%, but the authors did not feel that this was high enough to explain the inaccuracy. They proposed that the haemoglobin variant most probably possessed abnormal absorption spectra.15 This was subsequently verified in another patient with HbKöln.16 In addition, the authors of other case reports of patients with HbHammersmith and HbCheverly (both unstable haemoglobins) have suggested that inaccurately low readings have precluded the use of pulse oximetry.17 18 Furthermore, what is more interesting is that HbHammersmith exists almost entirely in the oxidized form but, in contrast, HbCheverly has reduced affinity for oxygen. This would seem to add more weight to the argument that these unstable haemoglobins have abnormal absorptive spectra.18 But how do we begin to explain why CO-oximetry is able to provide a measure of what is thought to be the true arterial oxygen saturation?
Modern pulse oximeters rely on the principle of spectrophotometry, the basis of which is the Beer–Lambert law. An alteration in the quantity of oxygen bound to haemoglobin results in a change in the absorption of light by the molecule. There are a number of differences between how pulse oximeters and CO-oximeters calculate and produce an arterial oxygen saturation measurement. For example, CO-oximeters measure haemoglobin oxygen saturation by the same spectrophotometric principle as the pulse oximeter, but they use an in vitro technique rather than an in vivo technique to evaluate haemoglobin: it is analysed in solution (the sample is haemolysed) and only for the moment it is taken from the patient.19 More importantly, CO-oximeters utilize more than two wavelengths for analysis, affording the ability to measure additional haemoglobins, such as carboxyhaemoglobin, methaemoglobin, and sulphhaemoglobin. They also function primarily in the visible range of the electromagnetic spectrum—avoiding the near infrared (940 nm) wavelength adopted in pulse oximeters.19 It could be argued that the dyshaemoglobin identified in our patient may have a similar absorption spectrum to normal oxyhaemoglobin in the visible range and only differs at the 940 nm wavelength; therefore, the CO-oximeter was able to actually measure it.
In the clinical situation, the CO-oximeter reassured us, but the drawback of this technique is that it was not able to indicate how much of the haemoglobin was abnormal. The ability to measure haemoglobin variants depends on the calibrated absorption spectra of the particular machine. In addition, we have no way of measuring the degree of oxygen affinity a dyshaemoglobin possesses. This may have significant clinical implications in the perioperative period in terms of oxygen delivery, placing the patient at risk of tissue hypoxia. In addition to CO-oximetry, modern arterial blood gas analysers are also able to provide a direct measurement of PaO2. In our patient, the normal PaO2 values again reassured us, but this could be short-sighted. A normal PaO2 does not signify normal oxygen transport capacity and in the context of a haemoglobinopathy with reduced oxygen affinity and where a large proportion of the patient’s haemoglobin is abnormal, this could again place the patient at risk of significant tissue hypoxia. Fortunately in this case, the dyshaemoglobin constituted only 7% of the total Hb.
To conclude, what have we learnt? We have revised some of the limitations of both pulse oximeters and CO-oximeters and have described how low pulse oximetry readings may be the result of a haemoglobinopathy. Through this, we have been informed of the presence of numerous haemoglobinopathies and their variable clinical significance. Ultimately, most haemoglobinopathies are benign and clinically inconsequential, but it must be remembered that some are associated with haemolysis, impaired oxygen binding, and secondary polycythaemia. In this case, we have been reassured that this potentially new haemoglobin variant has a benign pathology. One of the most pertinent issues when identifying a haemoglobinopathy is the counselling of the patient and subsequent family screening. Unfortunately, we were unable to commence this, but Bruns and colleagues20 have reasoned that pulse oximetry could easily be utilized to identify family members in these conditions where SpO2 values are low. However, knowledge of this patient’s previously undiagnosed haemoglobinopathy may avoid future needless investigations, such as chest radiographs, and unnecessary referral to a physician. It may also limit the use of oxygen therapy, and avoid the unwarranted occupation of a postoperative bed in a higher dependency setting, thus ultimately reducing patient distress.
|
References |
|---|
|
TopAbstractIntroductionCase reportResultsDiscussionReferences |
|---|
1 Moyle JTB. History of oxygen saturation monitoring. In: Pulse Oximetry—Moyle TB, ed. (2002) 2nd Edn. London: BMJ Books. 1–5.
2 The Association of Anaesthetists of Great Britain and Ireland Working Party (Chair RJS Birks). Recommendations for Standards of Monitoring During Anaesthesia and Recovery (2007) 4th Edn. Available from http://www.aagbi.org/publications/guidelines.htm. Accessed April 8, 2008.
3 National Haemoglobinopathy Reference Laboratory. Available from http://www.oxfordradcliffe.nhs.uk/forpatients/departments/labs/haemoglobinopathy/haemoglobinopathy.aspx. Accessed April 8, 2008.
4 HbVar. A database of human haemoglobin variants and thalassaemias. Available from http://globin.bx.psu.edu/hbvar/menu.html. Accessed April 8, 2008.
5 Moyle JTB. Limitations and morbidity. In: Pulse Oximetry—Moyle JTB, ed. (2002) 2nd Edn. London: BMJ Books. 140–52.
6 Jubran A. Pulse oximetry. Intensive Care Med (2004) 30:2017–20.[CrossRef][Web of Science][Medline]
7 Al-Shaikh B, Stacey S. Non-invasive monitoring. In: Essentials of Anaesthetic Equipment—Al-Shaikh B, Stacey S, eds. (1999) London: Churchill Livingstone. 86–100.
8 Moyle JTB, Davey A. Physiological monitoring: principles and non-invasive monitoring. In: Ward’s Anaesthetic Equipment—Ward C, ed. (1998) 4th Edn. London: Saunders. 263–78.
9 Mariette S, Leteurtre S, Lambilliotte A, Leclerc F. Pulse oximetry and genetic haemoglobinopathies. Intensive Care Med (2005) 31:1597.[CrossRef][Web of Science][Medline]
10 Higgs DR, Thein SL, Wood WG. Human haemoglobin. In: The Thalassaemia Syndromes—Weatherall DJ, Clegg JB, eds. (2001) 4th Edn. Oxford: Blackwell Science Ltd. 65–120.
11 Stephens A. Haemoglobinopathies: screening and diagnosis. The Biomed Scientist (2005) 139–44. Available from http://www.ibms.org/. Accessed April 8, 2008.
12 Old JM. Screening and genetic diagnosis of haemoglobinopathies. Scand J Clin Lab Invest (2007) 67:71–86.[CrossRef][Web of Science][Medline]
13 Moyle JTB. Dyshaemoglobins. In: Pulse Oximetry—Moyle JTB, ed. (2002) 2nd Edn. London: BMJ Books. 153–67.
14 Kuji A, Satoh Y, Kikuchi K, Satoh K, Joh S. The anaesthetic management of a patient with haemoglobin MIwate. Anesth Analg (2001) 93:1192–3.
15 Gottschalk A, Silverberg M. An unexpected finding with pulse oximetry in a patient with haemoglobin Köln. Anesthesiology (1994) 80:474–6.[Web of Science][Medline]
16 Katoh R, Miyake T, Arai T. Unexpectedly low pulse oximeter readings in a boy with unstable Haemoglobin Köln. Anesthesiology (1994) 80:472–4.[Web of Science][Medline]
17 Lang SA, Chang PC, Laxdal VA. Haemoglobin Hammersmith precludes monitoring with conventional pulse oximetry. Can J Anaesth (1994) 41:965–8.[Web of Science][Medline]
18 Hohl RJ, Sherburne AR, Feeley JE, Huisman THJ, Burns CP. Low pulse oximeter measured haemoglobin oxygen saturations with Haemoglobin Cheverly. Am J Hematol (1998) 59:181–4.[CrossRef][Medline]
19 Moyle JTB. Calibration. In: Pulse Oximetry—Moyle JTB, ed. (2002) 2nd Edn. London: BMJ Books. 35–51.
20 Bruns MB, Thet LA, Woodson RD, Schultz J, Hla KM. Haemoglobinopathy case finding by pulse oximetry. Am J Hematol (2003) 74:142–3.[CrossRef][Web of Science][Medline]
Read all E-letters
CiteULike 
Connotea 
Del.icio.us What's this?
E-letters:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||